To date, any brand wanting to embed Alexa – the industry leading voice assistant from Amazon – has had to use the Alexa keyword.
On 15th January 2021, all that changed with the announcement that Amazon will let companies build their own Voice Assistants via Alexa Custom Assistant. This enables companies to use the building blocks of the Alexa digital assistant to create and embed an own-brand version that has a customised wake-word and even a different voice.
This shift in strategy from Amazon could change the dynamic of the Voice Assistant in our everyday lives, with Alexa becoming one persona in a range of assistants that can interact with each-other to offer a broad range of services. This new offering will extend the software’s reach and help Amazon to fend off other rivals challenging for space in the smart home, smart car and smart workplace.
For the OEMs, it’ a game-changer. Until now, an OEM’s access to voice services from the major providers meant embedding another brand at the heart of their product (or adopting a diluted experience via a ‘works with’ connection to a nearby smart-speaker). In future, OEMs can build own-brand voice experiences to better differentiate their product, whilst delivering a full-featured, mature voice experience to their end-customers.
This is a busy playing field, with competing Voice Assistants from global giants and other companies providing mature white-label voice interfaces. However, Amazon’s offer of a consistent user experience across multiple collaborating assistants is disruptive, and it’s likely to further increase the pace of innovation in voice control and lead to better experiences for customers.
The announcement looks like good news for the consumer and great news for the OEM. But at what cost to Amazon? In the six years since Alexa came to market, Amazon’s strategy seems to have been to dominate the smart speaker category (and extend fast into the smart home), in order to drive its retail business. However, Amazon’s financials reveal that Amazon Web Service (AWS) makes up a third of the global cloud market (it’s twice the size of the nearest rival) and drives 57% of Amazon’s operating income from 17% of its revenue – and it’s growing fast. So the strategy for the Amazon Custom (voice) Assistant may be less about driving retail sales and more about AWS sales.
As smart speaker sales start to plateau, it’s certainly a smart move for Amazon to extend their reach across the voice ecosystem. They don’t have a native offer in the mobile phone market (unlike Apple and Google), PCs or even TVs, and the operating systems in those categories represent enormous barriers to entry for Amazon. But the IoT is open. This long tail market has a combined value that eclipses any individual category in the electronics market, but it’s characterised by a huge array of diverse products. How do you win in such a market? By having an enabling platform that OEMs can plug into easily across all categories to differentiate and diversity their product and service offerings.
Amazon’s latest move may cannibalise sales of Echo and spark a move away from the “works with” experience, but Amazon are likely to see a lot more adoption and usage via their Amazon Custom Assistant and since that “voice operating system” is hosted on AWS, the trade makes a whole lot of sense.
There’s still going to be a big drive towards intelligence at the edge, where voice commands are processed locally versus the Cloud for reasons of immediacy, privacy and cost efficiency. Many product categories will still benefit from a local voice interface (without needed to embed or connect to a voice assistant). However, Amazon’s Custom Assistant is disruptive. It really shifts the axis towards democratised voice capabilities and moves us closer to the point where our devices can interact with each other to deliver a more seamless, relevant and trusted experience.